Thermodynamic and artificial intelligence-based performance analysis of parabolic vacuum tube solar collector assisted greenhouse drying system

This study aims to increase energy efficiency by integrating renewable energy sources into agricultural drying processes. In the experiments carried out in Tokat climatic conditions, apple samples sliced with a thickness of 10 mm were used, and a total of 1573 data points were obtained with environmental parameters such as temperature, humidity, air velocity, and radiation. According to the experimental results, energy efficiency reached 7–33.4 %, exergy efficiency 4–7.4 %, and drying efficiency 61.5 %. Using these data, machine learning models were created with MLP, SVM, and M5P algorithms; the SVM algorithm provided the highest accuracy in exergy efficiency estimation with 0.0013 MAE and 0.0035 RMSE error rates. This study delivers a robust multivariate artificial intelligence modeling framework backed by actual experimental data, significantly advancing sustainable agricultural practices. It introduces a powerful decision support system designed for the intelligent control of parabolic trough solar collector systems, paving the way for more effective and environmentally conscious agricultural strategies.